JPH05313068A - Image input device - Google Patents

Abstract

PURPOSE:To provide an easily operable image input device which has a similar image quality to the whole object surfaces in a visual field and by which a focused image can be obtained and processing can be carried out properly on an optional object. CONSTITUTION:A lens driving device 102 moves a focusing position on an object surface by driving a lens 101. An adder 240 and an image memory 250 carry out a cumulative addition on mutual image signals photographed by a TV camera 103 while moving the focusing position by means of the lens driving device 102. A CPU 230 controls the lens driving device 102 by means of a characteristic of a driving speed function v(a) determined by the lens 101 and a characteristic of a correcting function w(a) calculated according to information on distances up to respective objects measured by a multi-areal distance measuring device 220 so that spatial frequency like characteristics of a plurality of object surface images in a visual field can be almost uniformized in a cumulative addition image.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image input device suitable for use in photographing an object image using an optical instrument such as an optical microscope.

[0002]

2. Description of the Related Art As is known, in an image input apparatus having an optical image forming system, the property of the input image depends on the characteristics of an optical image forming element such as a lens or an image pickup element. In general, as the aperture of the optical system increases, instead of improving the resolution,
The depth of focus tends to be shallow. Therefore, in order to increase the depth of focus in the conventional image input device, a method of controlling the size of the aperture with a diaphragm is used. However, it is difficult to obtain an image that matches the desired range of the object plane due to the characteristics of the optical system, and it is difficult to control the operation.

Therefore, when a target object exists in a range exceeding the depth of focus of the optical imaging system, a method of obtaining an image in which the range of an in-focus object plane (hereinafter, abbreviated as a focusing plane) is increased. As a method, a method disclosed in JP-A-60-68312 has been proposed. This is the distance between the sample and the lens barrel when focusing on the most convex part of the sample in the field of view in the optical microscope, and the sample when focusing on the most concave part of the sample. With respect to the distance between the lens barrels, at least one of the sample and the lens barrels is configured to scan in the optical axis direction of the optical microscope from one distance to the other distance within the photographing exposure time. According to this method, since the light amount is sequentially integrated on the film, only the in-focus portion is exposed accordingly, and as a result, the in-focus image is obtained as a whole. ..

However, in the method disclosed in this publication,
Only the convex and concave parts of the sample are taken into consideration, and it is assumed that the out-of-focus image has a constant background intensity. You can't get a picture.

[0005]

As described above, in the conventional image input apparatus having the optical imaging distortion, the aperture control such as the diaphragm is used to adjust the depth of focus.
It was difficult to achieve the desired depth of focus and difficult to operate. Further, with the image input method for increasing the depth of focus that has been conventionally proposed, it is not possible to obtain an image having the same image quality for all object planes within the field of view.

The present invention has been made in view of the above points, and has the same image quality for all object planes in the field of view, and an in-focus image can be obtained, and an arbitrary target object can be obtained. It is an object of the present invention to provide an image input device that can be adaptively processed and that is easy to operate.

[0007]

In order to achieve the above object, an image input device of the present invention converts an optical image forming system and an image of an object formed by the optical image forming system into an electric signal. Image pickup means, a focus plane moving means for moving the focus position on the object plane, and a cumulative addition between the image signals obtained by the image pickup means while moving the focus position by the focus plane moving means. The accumulative addition means composed of an addition means and an image memory, and the spatial frequency characteristics of the images of a plurality of object planes within the field of view become the most uniform in the accumulated addition image obtained by the cumulative addition means. As described above, it is characterized by comprising the focusing surface control means for driving and controlling the focusing surface moving means.

[0008]

That is, in the image input apparatus according to the present invention, the focusing surface control means adds the cumulative addition between the image signals obtained by the image pickup means and the cumulative addition means formed by the image memory. In the added image, the focus plane moving means for moving the focus position on the object plane is driven and controlled so that the spatial frequency characteristics of the images of the plurality of object planes within the field of view become the most uniform. I have to.

That is, when the input images are cumulatively added while moving the focus position (focus plane) over a predetermined range,
Image information focused on the target object surface within the range is included in the cumulative addition image. Therefore, even if a plurality of target object planes exist at different positions, it is possible to capture the focus information of all the target object planes by appropriately setting the moving range of the focus plane. In that case, not only the focus information but also the out-of-focus information, that is, the blurred image is added to the image of the target object surface in the cumulative addition image. However, in reality, since the focus information has a dominant influence, the influence of image deterioration due to addition of the non-focus information is small. Further, the image input device according to the present invention has a configuration in which the focusing surface is driven and controlled so that the spatial frequency characteristics of the images of the respective target object planes in the cumulative addition image become the most uniform, whereby the cumulative addition is performed. It has a function of correcting the image quality of each target object surface in the image, that is, the variation in focusability.

[0010]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing the configuration of the first exemplary embodiment of the present invention. The image input device according to the present embodiment is roughly composed of an image pickup device 100, an image processor 200, and a TV monitor 30.
0, control unit 400.

Of these, the image pickup apparatus 100 includes a lens 10
1, a lens driving device 102, and a TV camera 103. The lens driving device 102 includes the image processor 20.
The focusing surface is set at a predetermined position in response to a command signal from the CPU 230 within 0, and the lens 1
An encoder (not shown) for detecting the position of a focus lens (not shown) included in 01 is built in, and information on the position of the focusing surface is sent to the CPU 230. An image signal obtained by being imaged by the lens 101 and imaged by the TV camera 103 is obtained by the image processor 2
It is converted into a digital signal by the A / D converter 210 in 00.

In this configuration, preprocessing for measuring the distance from the TV camera 103 to each object plane in the field of view, and focusing on all object planes in the field of view based on the distance information obtained by the preprocessing. A two-stage operation is performed, which is the main processing for synthesizing the images.

First, the configuration related to preprocessing will be described. Input from the image pickup apparatus 100, and A / D converter 210
The image signal digitally converted by is input to the multi-area distance measuring device 220. The multi-area distance measuring device 220 is configured as shown in FIG. 2A, and the distance to each object plane in the field of view is measured by measuring the contrast of each small area obtained by dividing the image. It is like this. That is, when the digital image signal is input to the multi-area distance measuring device 220, first, only a predetermined frequency band is extracted by the bandpass filter (BPF) 221, and then the signal is squared by the squarer 222. Cumulative addition is performed by the adder 223, the latch 224, and the memory 225.

As shown in FIG. 2B, the above operation is address-controlled so that all the pixels in each divided image are cumulatively added, and finally the memory 225 stores a predetermined value for each divided image. The cumulative value (= contrast) of the power in the frequency band of the frequency is recorded.

The above operation is repeated by the lens driving device 102 as the in-focus surface is moved little by little, and the memory 225 stores the characteristic () of the contrast at the in-focus position as shown in FIG. = Contrast curve) is recorded.

Then, in the CPU 230, the position of the vertex of the contrast curve of each divided image and the lens driving device 1
The distance of the object plane imaged in each divided image is measured from the data value corresponding to the in-focus position sent from the unillustrated encoder in 02.

Next, in this processing, the operation with the following configuration is performed. That is, based on the distance information to each object plane obtained by the preprocessing as described above, the lens driving device 1
The image signal input by the image pickup apparatus 100 after being focused on a predetermined object plane by 02 is converted into a digital signal by the A / D converter 210, and then cumulatively added in image units by the adder 240 and the image memory 250. To be done. The image signal recorded in the image memory 250 after being focused on a plurality of predetermined object planes and input / added is subjected to filtering for emphasizing a predetermined middle-high frequency spatial frequency by the spatial filter 260, The output is the D / A converter 2
It is converted into an analog video signal by 70 and displayed on the TV monitor 300.

The CPU 230 controls the above operations, and the control unit 400 controls the CPU 2
It is connected to 30 and is configured to transmit an operator's command to the device as a man-machine interface.

Next, the method of setting the focal plane in this processing will be described. The lens 101 is usually composed of a plurality of lens groups, but the in-focus position is set by driving a lens, which is called a focus lens, in the optical axis direction. FIG. 3A shows an example of the relationship between the drive amount of the focusing lens and the focus position. In the figure, the horizontal axis represents the encoder address value a indicating the lens drive amount, and the vertical axis represents the focus position h (a) with respect to the encoder address value a.

On the other hand, the depth of focus (depth of field) d (a) for the in-focus position h (a) is approximately represented by the following equation (1). That is, d (a) = d ₂ (a) −d ₁ (a) (1)

[0021]

[Equation 1]

F: focal length, p: coefficient determined by the resolution of the image sensor, k: coefficient determined by the optical system. D ₁ (a) expressed by the above equation (2) represents a near point of the depth of focus, and d ₂ (a) expressed by the above equation (3) represents a far point of the depth of focus.

Using these relationships, the lens driving speed function v (a) is defined as shown in the following equation (4). That is, v (a) = {c / h (a)} d (a) (4)

That is, the above equation (4) takes the reciprocal of the focusing position h (a) so that the focusing position moves at a constant speed on the object side with respect to the driving of the lens, and each focusing surface The image is corrected so that the image is input in proportion to the depth of focus d (a). Therefore, the driving speed function v of this lens
By driving the lens based on (a), it is possible to prevent sparseness and denseness from occurring in the set distance of the focusing surface in the object surface, and in the object surface with a large depth of focus, it becomes unnecessary. It is possible to prevent inputting images at fine intervals. The relationship between the focus position h (a), the depth of focus d (a), and the lens driving speed function v (a) is shown in FIG. 3 (B).

Next, a lens driving method in consideration of the arrangement of objects in the visual field will be described. As shown as an example in (A) of FIG. 4, it is assumed that the object in the visual field is at positions A _{1 to} A ₃ . At this time, based on the driving speed function v (a) of the lens, even if the image is input and accumulated while moving the in-focus position from the position A ₁ to the position A ₃ , the image quality of each object is uniform. I won't. This is because, as shown in the above formulas (1) to (3), the depth of focus tends to become deeper as the focus position is set on the object plane farther from the TV camera 103, and the focus plane for each target object tends to be deeper. This is because there is a difference in the deterioration state of another object when the two are combined. That is, the image of the object at the farthest position A ₃ is less deteriorated even if the object plane relatively close to the camera 103 is focused, whereas the image of the object at the closest position A ₁ is The focus surface deteriorates rapidly as the distance from the focus surface increases, and when the focus position is set to the position A ₃ , the structure becomes almost invisible so that the structure is blurred. Therefore, in the image after the cumulative addition, the object / image at the position A ₁ is expected to be deteriorated more than the image of the object at the position A ₃ .

Therefore, a correction function w (a) that makes the deterioration of the image of each object in the cumulative addition image uniform is set as follows. First, the average value δ _j (j = 1, 2, 3) of the diameter of the circle of confusion when the weighting factors C ₁ , C ₂ , C ₃ are applied to the object planes A ₁ , A ₂ , A ₃ respectively, and the cumulative The weighting coefficient is calculated so that the average values δ ₁ and δ _{2 of} A ₁ having the largest deterioration and A ₂ having the smallest deterioration in the added image are closest to each other. Actually, the calculation as shown in the following equation (5) is performed.

[0027]

[Equation 2]

Where t _{ij is} the distance between A _i and A _j (i, j = 1,2,3) δ _j (t _ij ): is the position of A _j when the focal plane is at the position of A _i The diameter of the circle of confusion. Then, f expressed by the following equation (6) is a constraint condition.

[0029]

[Equation 3] Apply Lagrange's undetermined multiplier method to minimize under.

[0030]

[Equation 4] However,

[0031]

[Equation 5] Here, ψ defined by the above equation (7) is a weighting coefficient C
Partial differentiation is performed with _j (j = 1, 2, 3), and 0 is set.

[0032]

[Equation 6] From these equations (8) and (9), a quaternary simultaneous linear equation is established, and by solving it, the weighting coefficient C _j (j = 1,
2, 3) is required.

Then, the weighting coefficient C _j is interpolated by a multi-dimensional curve to obtain the correction function w (a). Figure 4
In (B) of, the weighting factors C ₁ , C ₂ , C ₃ and the correction function w
An example of (a) is shown.

Finally, the lens driving speed function v
_w (a) is the drive speed function v obtained by the above equation (4)
It is obtained as a function obtained by multiplying (a) by the correction function w (a). _{That, v w (a) = v} (a) · w (a) = {c / h (a)} d (a) w (a) ... (10)

In the device configuration, the driving speed function v (a) is determined by the lens 101, so the characteristics of this driving speed function v (a) are recorded in advance in the ROM or the like in the CPU 230, and , For the correction function w (a),
It is calculated in the CPU 230 based on the distance information to each object measured by the multi-area distance measuring device 220.

The lens driving device 102 may be configured from the beginning so as to drive the lens based on the driving speed function v (a). Further, several correction functions w (a) are stored in the ROM provided in the image processor 200, and the distance information to each object is input, and the correction function w (a) of the closest condition is called. It may be configured.

The spatial filter 260 is used for emphasizing the mid-high spatial frequencies in order to make the image in which the deteriorated image is cumulatively added more sharp.

As described above, in the present embodiment, in the image input device using the camera lens, it is possible to synthesize focused images with the same image quality for objects within the field of view and at different distances. In addition, it becomes possible to perform the optimal processing adaptively to the conditions of the arbitrary target object.

Next, a second embodiment of the present invention will be described.
The configuration of the second embodiment of the present invention is shown in the drawing. The configuration is roughly divided into a microscope device 500, a TV camera 600, a stage drive device driver 700, an image processor 200, a TV monitor 300, and a control unit 400. Of these, the image processor 200, the TV monitor 300, and the control unit 400 are the same as those in the first embodiment, and therefore their explanations are omitted.

The microscope apparatus 500 is provided with a stage drive unit 501 and a stage position encoder 502, and the stage drive unit driver 700 operated based on a command from the CPU 230 in the image processor 200 is used to drive the microscope unit 500. Predetermined operation control is performed on the stage in the optical axis direction. TV camera 600
Is installed on the barrel of the microscope apparatus 500, and a microscope image is captured.

Also in the second embodiment, the operation is performed in two stages of preprocessing and main processing. The pre-processing is the above-mentioned first
Similar to the embodiment described above, the distances to the respective target object planes within the field of view are measured by measuring the multi-area contrast while moving the stage in the optical axis direction by the stage driving device 501.

Also in this processing, as in the case of the first embodiment, the in-focus image is set at a plurality of predetermined positions based on the distances to the target object surfaces obtained by the pre-processing, and the input image is It is cumulatively added.

The method of setting the focal plane in the second embodiment will be described below. To simplify the explanation,
Assume an object having a layered structure composed of a plurality of object planes as shown in FIG.

In the method of the second embodiment, the in-focus plane is set so as to include the range of the vertical distance dr for each object plane. That is, as shown in FIG. 6, for the object plane 1 in the position of _{z 1 z 1u = z 1 -d} r from z _1h = z
₁ + d _r range, similarly for the object plane 2 at the z ₂ position z _2u = z ₂ -d _r to z _2h = z ₂ + d _r ,
For the object plane 3 at that position, the range from z _3u = z ₃ −d _r to z _3h = z ₃ + d _r is included. Therefore, for all the object planes, according to FIG. 6, the images may be input while moving the focusing plane from z _2u to z _3h and from z _1u to z _1h , and cumulative addition may be performed.

The operation of the above method will be described below. First, the input image of the target object surface t when the object surface s is focused is expressed by the following equation (11). That is, g _{t (x, y; t, s)} = h _{(x, y; t, s)} * f _{1 (x, y)} + n _{(x, y)} (11) where h _{(x, y; t, s)} : transfer function (PSF) of the target surface t when focused on the object surface s, ft _{(x, y)} : original image of the target surface t, n _{(x, y)} : additively Noise added, g _{t (x, y; t, s)} : image of the observed target surface t, *: operator representing convolution. In the case of a microscope optical system, PSF is

[0046]

[Equation 7] You may think that it depends on. Therefore, in the following, (t, s)
Will be rewritten as z.

Now, consider inputting and adding images while discretely changing the focal plane. The image of the target surface t in the added image can be expressed by the following equation (12) derived based on the equation (11).

[0048]

[Equation 8] However,

[0049]

[Equation 9]

Here, h _{a (x, y)} is a function component h _{r (x, y} that contributes to the image formation of the original image f _{t (x, y)} , as shown in the following equation (13) _{. It} is considered to be the sum of _y) and a functional component h _{n (x, y)} with large deterioration that transmits almost only noise. h _{a (x, y)} = h _{r (x, y)} + h _{n (x, y)} (13) where

[0051]

[Equation 10] r: A number representing the defocus range that contributes to image formation.

[0052]

[Equation 11] In this case, g _{ta (x, y)} is expressed by the following equation (16).

[0053]

[Equation 12]

However, M = N- (2b + 1): the number of input images in which noise is dominant, and the spatial average value of _Ct : _{ft (x, y)} .

In the above equation (16), the first term is interpreted as an in-focus image, the third term is interpreted as a bias term in which the image structure disappears due to large deterioration, and the fourth term is interpreted as noise. .. Therefore, if the focus plane is set in a range in which the function components h _{r (x, y)} contributing to image formation are completely included in all object planes, the above (1
Expression 6) is similarly established. That is, the deterioration state of each target surface in the image after addition becomes uniform. Further, if the spatial average value C _t does not change depending on the target surface, the uniformity including S / N is established. The range dr that contributes to image _formation is
It may be appropriately determined depending on the characteristics of the optical imaging system, but as an example, a method of using the defocus amount of the optical imaging system or the defocus amount at which 0 point first appears in the OTF can be considered.

In the case of an optical image-forming system having a large NA such as an objective lens in a microscope, the depth of focus is relatively shallow, and the blurring due to defocusing is almost the same before and after the in-focus surface, and in-focus. The optical characteristics such as the depth of focus and the OTF do not change with respect to the position of the surface. In the image input device having such an optical system, the second embodiment adaptively processes a target object having an arbitrary structure to obtain the frequency characteristics and S / S of the image of each target object surface. N
It is possible to provide a device for inputting an image such that the image becomes uniform.

[0057]

As described in detail above, according to the present invention,
Has similar image quality for all object planes in the field of view,
In addition, it is possible to provide a practically useful image input device that can obtain a focused image, can adaptively process an arbitrary target object, and is easy to operate.

[Brief description of drawings]

FIG. 1 is a block configuration diagram of an image input device according to a first embodiment of the present invention.

2A is a block diagram showing details of the multi-area distance measuring device in FIG. 1, FIG. 2B is a diagram showing divided images, and FIG.
Is a graph of a contrast curve recorded in the memory in FIG.

FIG. 3A is a graph showing a relationship between a driving amount of a focusing lens and a focusing position, and FIG. 3B is a focusing position h (a), a depth of focus d (a), and a lens driving speed function. It is a graph which shows the relationship of v (a).

FIG. 4A is a diagram showing a positional relationship of objects in a visual field, and FIG. 4B is a graph showing an example of weighting coefficients C ₁ , C ₂ , C ₃ and a correction function w (a). ..

FIG. 5 is a block configuration diagram of an image input device according to a second embodiment of the present invention.

FIG. 6 is a diagram showing an object having a layered structure composed of a plurality of object planes.

[Explanation of symbols]

100 ... Imaging device, 200 ... Image processor, 220 ...
Multi-area distance measuring device, 230 ... CPU, 240 ... Adder, 250 ... Image memory, 300 ... TV monitor, 400
Control unit, 500 Microscope device, 600
... TV camera, 700 ... Stage drive device driver.

Claims (1)

[Claims] 1. An optical imaging system, an imaging means for converting an image of an object formed by the optical imaging system into an electric signal, and a focusing plane movement for moving a focusing position on an object plane. Means, an addition means for performing cumulative addition between the image signals obtained by the image pickup means while moving the focus position by the focus plane moving means, and a cumulative addition means made up of an image memory; Focusing surface control means for driving and controlling the focusing surface moving means so that the spatial frequency characteristic of the image of the object surface of the object becomes the most uniform in the cumulative addition image obtained by the cumulative adding means. An image input device comprising.